JP4304155B2 - Automatic frequency control receiver - Google Patents

Abstract

The invention relates to a receiver of signals [S] received from a wireless network, said receiver working at a so-called reference oscillation frequency controlled by a so-called reference value [Vref]. Said receiver includes demodulation means [DEMO] for demodulating the received signal [S], means [EST] of estimating a mean value [MV] of the demodulated signal [SD], means [COR] of correcting the mean value [MV] of the demodulated signal [SD] to the reference value [Vref], decision means [DEC] for determining the binary values adopted by the received signal [S]. According to the invention, the estimation means [EST] include first means [ESTA] of fast extraction of a first mean value [MVA] of the demodulated signal [SD] used in decision means [DEC] during a first time period and second means [ESTB] of slow extraction of a second mean value [MVB] of the demodulated signal [SD] used in correction means [COR] and, during a second time period, in decision means [DEC].

Description

  The present invention relates to wireless receivers, and more particularly to removing frequency differences normally found in signals received by a wireless receiver between the local frequency of the receiver and the local frequency of the received signal. For example, in the BlueTooth standard, the receiver must be able to handle large frequency differences. In general, in a wireless system that includes a transmitter that communicates with a receiver, there is a difference between the frequency of the transmitter and the frequency at which the local oscillator of the receiver, called the reference frequency, operates. This frequency difference is expressed as a voltage difference at the output of the modulator of the receiver. A method of estimating the voltage difference from the analysis of the demodulated signal over a given period and then removing the voltage difference from the signal, or this voltage to adjust the local oscillator in an automatic frequency control (AFC) loop There already exists a way to use the difference. Therefore, in the conventional example, a radio receiver operating at a so-called reference oscillation frequency controlled by a so-called reference value includes a demodulating means for demodulating the received signal, a means for estimating an average value of the demodulated signal, and an average value of the demodulated signal. Means for correcting the reference value to a reference value, and judgment means for determining a binary value acquired by the received signal.

  Therefore, the function of correcting the average value of the demodulated signal by the automatic frequency control loop is known. In this loop, the estimation means for estimating the voltage difference can be configured, for example, by passing a demodulated signal through a low-pass filter having a narrow passband that removes useful signals and preserves continuous components.

  The present invention relates to the following considerations.

  The selection of the pass band of the low pass filter is a compromise, and this method with a narrow pass band is slow to estimate the continuous components. If the passband is widened to increase the speed of operation, the estimation of successive components may be corrupted by the component of the signal passing through the filter. The disadvantage of the slow method is that it takes a long part of the signal to estimate the voltage difference before receiving useful information, which is a loss of time or accuracy in detection for each start of the reception range. Is to cause. A faster, but worse, estimation disadvantage is that errors are introduced into the received signal by the method of estimating the voltage difference.

  One object of the present invention is to make it possible to obtain a receiver for signals received through a wireless network, which generally does not have the disadvantages of the prior art described above, depending on the reception range.

  For this purpose, the radio receiver, characterized by the invention, based on the introductory paragraph, quickly extracts the first average value of the demodulated signal used in the decision means during the first period. The estimation means includes a first means, and a second means that is used in the correction means and extracts a second average value of the demodulated signal used in the determination means at a low speed during the second period. .

  The two types of means for estimating the average value are complementary, and even slow systems where a long portion of the signal is required to estimate the average value can be introduced into the signal by incorrect estimation of the signal average value. It also makes it possible not to have. Therefore, the obtained signal receiving circuit combines the low-speed estimation and the high-speed estimation of the frequency so that the accurate correction of the frequency and the sensitive receiver can be realized simultaneously from the start of the reception range.

  In a particular embodiment of the invention, the means for correcting the average value of the demodulated signal uses a frequency correction loop. The principle of automatic frequency control is known as the principle for analog data and digital data.

  In a useful embodiment of the invention, the first estimation means evaluates the minimum and maximum values of the received signal, thereby estimating the average value of the signal as an intermediate value between these two values. including. Such an estimation is very quick. However, the estimation suffers relative inaccuracies (eg, when a series of consecutive 1s and 0s (zeros) are present in the signal). This estimation can be used for the entire signal, but the accuracy of the decision means can be influenced thereby.

  For this reason, in the preferred embodiment of the present invention, when the received signal is composed of the synchronization and control unit and the subsequent data unit, the first period does not exceed the period necessary for reception of the synchronization and control unit.

  The present invention can be implemented in any receiver operating in a wireless network where large frequency differences can occur. Therefore, it relates to the BlueTooth technology, the Digital European Cordless Telecommunications (DECT) technology, and any other technology that exhibits the characteristics described above.

  The principle according to the invention can be used for digital data or analog data as well.

  The invention also relates to an integrated circuit with a receiver according to the invention.

  The invention also relates to a method for receiving and processing signals received from a wireless network according to the invention.

  The following description relates to reference numerals. In all of the figures, similar components are represented by the same letter. Several similar components may appear in a single figure. In this case, numbers or subscripts are added to the letter reference signs to distinguish similar components. Numbers or subscripts may be omitted for reasons of convenience. This applies to the description and the claims.

  The following text is presented to enable any person skilled in the art to implement or use the present invention. The text of this specification is provided in terms of patent applications and their requirements. Various modifications to the preferred embodiment will be apparent to those skilled in the art and the generic principles of the invention disclosed herein may be applied to other implementations.

  Accordingly, the present invention is not considered limited to the described embodiments, but has the broadest scope in accordance with the principles and characteristics set forth below.

  FIG. 1 shows a block diagram of a receiver according to the invention. This figure shows the means for performing the various functions necessary to implement the present invention. These functions are steps in the receiving method according to the present invention. The receiver according to the invention receives the signal S from the wireless network. The receiver has a receiving means (not shown), and may be an antenna, for example. The receiver operates at a so-called reference oscillation frequency, and the oscillation frequency is controlled by a so-called reference value Vref. This reference value is usually a voltage. The receiver according to the present invention comprises a demodulating means DEMO for demodulating the received signal S, a means EST for estimating the average value MV of the demodulated signal SD, and a means for correcting the average value MV of the modulated signal SD to a value at the reference value Vref. COR and determination means DEC for determining a binary value acquired in the received signal S. According to the invention, the means EST for estimating the average value of the received signal first means for rapidly extracting the first average value MVA of the modulation signal SD used in the decision means DEC during the first period. ESTA and second means ESTB for extracting at low speed the average value MVB of the modulation signal SD used in the correction means COR and used in the determination means DEC during the second period. The data handled by the receiver according to the present invention may be digital or analog. In the following, application to the analog processing of digital data is presented. In that case, the means for estimating the average value is a method intended to process analog data. Digital means for estimating digital data can also be used without compromising the principles of the present invention.

  In a particular embodiment, and according to what is shown in FIG. 1, the means COR for correcting the average value MV of the demodulated signal SD is looped to the demodulator DEMO, thereby forming what is commonly called a frequency correction loop. . Such a loop is generally known by the term automatic frequency control. FIG. 2 shows a block diagram of the general principle of an automatic frequency control loop. The function of such a frequency control loop is double. This keeps the frequency fIF constant regardless of the frequency difference in the received signal. It also compensates for any changes in method and temperature by automatically correcting the frequency fIF to the center frequency of the modulator. This is accomplished by looping the information at the average frequency fIF to the local oscillator VCO of the receiver, so that the local frequency fVCO is automatically adjusted. In order to extract the average value of the signal frequency, the output signal of the modulator DEMO is filtered in a low-pass filter FIL. Therefore, this filter FIL constitutes means ESTB for low-speed estimation of the average value MVB of the modulation signal SD. An example of such a filter is shown in FIG. The average value MVB is compared with the reference value Vref. This generates, for example, an error signal. This information is amplified and applied to the control input of the local oscillator VCO. This leads to changes in the frequency of the VCO that help cancel out the frequency difference between the signal and local oscillation.

More precisely, the demodulator is assumed to have a linear response in the form

Here, if fIF is equal to fC and Kd is the gain of the demodulator, Vc is the output voltage. Thus, the voltage SD is available at the demodulator output. The demodulator DEMO is followed by means ESTB for estimating the average value MVB of the signal SD. These estimation means ESTB advantageously constitute the low-pass filtering means FIL, as can be seen above. The obtained average value MVB is proportional to the frequency difference between the frequencies fIF and fc. The average value MVB is then compared with a reference value, usually a voltage Vref, amplified by the gain A amplification means AMP and fed to the local oscillator VCO of the receiver. This is summarized in the following equation:

The output signal of the amplifier AMP is supplied to a local oscillator VCO that functions at a frequency fIF. Therefore,

fref is a frequency when Vmod = Vref and Km is a demodulation gain. Thus, each difference in signal frequency corrects the current frequency fIF, which is the same value but in the opposite direction. And

  Second, the response to changes in the frequency RF of the system can be a difference in the carrier frequency or the modulation itself.

Using the previous equation, the following equation is finally obtained.

The equation is simplified if Vc = Vref.

  Ideally, the system is configured such that fc = fref. In other words, if fIF = fref, this means that the output voltage Vc = Vref. This cannot be the case, for example, due to method variations. If the frequency difference of the signal ΔfRF is zero and the gain of the loop kmAFKd is quite high, the equation shows that the loop repositions the frequency fIF to fIF = fc.

  If there is a frequency difference in the received signal, the voltage difference is attenuated by the factor 1 + KmAFKd, so the loop keeps the frequency fIF close to fc.

  The loop must compensate for the slow drift of the frequency of the received signal, but must not react to the modulation. This is achieved by optimization of the loop filter.

By selecting a first order filter with a transfer function,

For fc = fref, the result is

  Here, Group = KmAF (0) Kd is a static loop gain.

  For a received signal that changes rapidly such that τp / (1 + Group) >> 1, the loop does not respond and therefore the modulation is transmitted without attenuation in the loop. In order to improve the sensitivity of the voltage present in the loop filter capacitor, the loop amplifier must have a high input impedance. Its output controls the tuning of the local oscillator VCO. For example, the amplifier is realized by a transconductance amplifier with a return according to FIG.

The loop amplifier can be classified into a gain term Gv and a return gain term β.

The closed loop gain is

  For example, in this application approximately 20 gains are required. Since the transconductance is a single stage amplifier, the gain is not very large and the target gain is achieved with a much lower return (slightly greater than a single Gvβ). Here we looked at the general function of an automatic frequency control loop. In application to a receiver operating in a wireless network, for example a Bluetooth (BlueTooth) receiver, for example, a decision means DEC for determining whether the signal corresponds to 1 (1) or zero (0), In other words, the determination means DEC that tries to determine the binary value of the received signal is necessary. The automatic frequency control loop and decision means DEC require extraction of the average value of the signal, but what is required for this extraction is different. For the decision means DEC, it must be extracted particularly quickly during the start of the reception range of the receiver. The reception period start period generally corresponds to an access code, signal synchronization, and error control signal. On the other hand, for frequency control loops, slow extraction is permanently required to prevent attenuation or cancellation of modulation by the frequency control loop. During the start of the reception range, the extraction of successive components is preferably performed by a first means ESTA that estimates the average value. In this way, the loop can obtain a low gain (eg 6). The period of extraction performed by the various means EST for estimating the average value MV is selected according to the characteristics of the signal to be received by the receiver. According to FIG. 4, these periods are controlled by the control means, which are switches controlled, for example, by the time diagram disclosed in FIG.

  FIG. 4 illustrates a useful embodiment of the present invention. The means presented in FIG. 1 are again found in more detail and in terms of analog processing of the data.

  According to FIG. 4, the receiver receives the signal S in its demodulator DEMO. In order for the system to receive data, the demodulator activates switch R-ON. Once demodulated, the demodulated signal SD is supplied as an input to the two means ESTA and ESTB for estimating the mean values MVA and MVB. The first means ESTA is fast, for example performing an evaluation of the minimum and maximum values of the signal and then taking an intermediate value between these two values as the average value of the demodulated signal. One implementation of such an evaluation is depicted in FIG. The evaluation of signal extrema and the calculation of intermediate values, as described in particular in FIG. 4, are well known from the prior art and are therefore not described in more detail here. According to FIG. 4, the second estimator ESTB comprises, for example, low-pass filters R2, C1. The value of the capacitance C1 is selected from a compromise in the RC filter time constant called ESTB. Since this value is usually high, the capacitor is external. To obtain a good estimate of the average value of the demodulator output signal, the time constant must be quite long, but if the reference input of the decision means is switched to the output of the slow RC filter, the output of the filter is In order to ensure that the end value is already approaching the end value at the end of the access code, the time constant must be quite short. Surprisingly, the invention makes it possible to have only one external capacitor in the circuit, which means saving space and manufacturing time in the receiver. This single outer capacity is permanently used for the frequency control loop, and at intervals, used for decision means.

  As shown in FIG. 2, the correction means has an amplifier AFC connected to the input of the local oscillator VCO and is placed in one loop with the demodulator DEMO as a whole. The decision means DEC has an amplifying element SLI that fixes the value of the output of 1 or 0 (zero) depending on the values at the two inputs, the two inputs being the first from one of the estimation means ESTA and ESTB. The average value of the incoming signal is received, and the second is the demodulated signal SD itself. Advantageously, this amplifying element SLI also deactivates the switch signal R-On activated by the demodulator DEMO when there is no signal to process. According to the preferred embodiment, the various diagrams are activated by the time diagram presented in FIG.

  By activating the switch S-EN in the first column of the time diagram in FIG. 5 before the start of the reception range in FIG. 5, ie, at the receiver power-up, the external capacitor C1 is pre-charged to the reference voltage. Charged. When the receiver is powered up, the local oscillator VCO and loop PLL are also started. At the start of the reception range, signal detection is required for all elements of the receiver by the signal R-ON triggered by the demodulator DEMO. Thereafter, the reception REC is started, and the data arrives in the form as shown in FIG. 5, for example. These data composing the received signal are an access code consisting of several synchronization and error check elements PR, SYN, TR, followed by specific data that is specific to the next communication processed by the receiver. It is formed by one PAYL packet. Thereafter, the switch R-ON switches on the external capacitor C1 to the output of the demodulator DEMO, and the average value MVB of the signal is extracted by the second estimation means ESTB. A relatively long time constant is used. When there is a frequency difference, the average value MVB of the signal is different from the reference voltage Vref. This information is fed to the input of the local oscillator VCO to correct the receiver frequency. This forms what is known as automatic frequency control. During this time, the receiver must determine whether the received bit is zero or one. This must also be done by comparison with the average value MV of the signal. The average value MVB supplied by the estimation means ESTB does not converge fast enough to the actual average value and cannot be used at the beginning of the reception range. Therefore, another circuit is used. This circuit has an estimation means ESTA and determines an average value from the detection of the peak. This estimation means ESTA is reinitialized by the activation of the switch SR. Then, the peak detection and the peak value MIMA are evaluated during the time constant shown in FIG. During this evaluation, switch S2 is activated and switch S3 is opened, so that the mean value of the signal is available at the output of the estimation means ESTA and at the reference input of the decision element SLI. Next, by switching the switch S2 to the open position and closing the switch S3, the reference input to the determination means DEC is switched to the average value MVB measured by the estimation means ESTB. In this way, the frequency control loop and the decision means use the same average value MVB of the signal. The time constant is selected such that switching occurs before a data packet PAYL containing data specific to the next communication processed by the receiver arrives.

The various elements in FIG. 4 were simulated with the values proposed in the table below. This simulation is proposed in FIG.

  These values are provided for purposes of illustration and do not exclude other values or ranges of values. For example, if the time constant for data acquisition and adjustment of the average value to the reference value can be increased for a particular application, the external capacitance C1 can also be increased. The simulation was performed for a difference in carrier frequency of, for example, around 100 kHz (200 kHz in the example of FIG. 6). In such a simulation, for example, operations performed in the time diagram in FIG. 5 are shown. The loop PLL and frequency control loop are first activated. The frequency control loop is initialized with the same reference voltage Vref as used in the local oscillator, for example 1.45V. After the loop PLL is established at the required frequency, the receiver is activated in 200 μs according to the diagram of FIG. The data reaches the curve Demod-OUT at 225 μs. In FIG. 6, these data have a difference with respect to the reference voltage Vref charged by the external capacitor and indicated by the curve Cext. In this simulation, the advantages of the present invention show the effect of a frequency control loop that restores data and thereby restores the curve Demod-OUT demodulated near Vref represented by the curve Cext (VinAFC). By the detection of the peak of the curve Demod-OUT by the extraction means, the average value used by the judgment means indicated by the curve DC-Slicer is first supplied, which is the data up to approximately 350 μs in FIG. 5 and FIG. Operates on packet access codes and header packets. At this point, the frequency control loop is correcting the initial difference. Due to the peak detection ESTA, the level of the continuous component MVA extracted by the extraction means approaches the average level MVB indicated by the external capacitor C1. At that time, the reference input of the judging means is switched to the external capacitor C1.

  Although the present invention has been described in accordance with the presented embodiments, those skilled in the art will readily recognize that variations to the presented embodiments exist and are within the spirit and scope of the invention. Will do. Accordingly, many modifications may be made by one skilled in the art without departing from the spirit and scope as defined by the following claims.

The invention will be further described with reference to the example embodiments shown in the drawings, but the invention is not limited.
FIG. 1 is a block diagram of a receiver according to the present invention. FIG. 2 is a block diagram of the general principle of an automatic frequency control loop. FIG. 3 is an example of an embodiment of an amplifier implemented in a frequency control loop. 3 is a functional diagram of a receiver according to a useful embodiment according to the present invention. 2 is a time diagram for controlling elements of the present invention according to a preferred embodiment of the present invention. FIG. 5 shows the result of frequency correction obtained with the implementation of the present invention by a set of curves representing the voltage at several points in the circuit of FIG.

Claims (7)

A receiver for signals received in a wireless network, wherein the receiver operates at a so-called reference oscillation frequency, the oscillation frequency being controlled by a so-called reference value,
Demodulation means for demodulating the received signal;
Means for estimating an average value of the demodulated signal;
Means for correcting an average value of the demodulated signal to a value at a reference value;
Determining means for determining a binary value obtained by the received signal,
The estimation means includes
First means for quickly extracting a first average value of the demodulated signal used in the determining means during a first period;
Second means for extracting at low speed a second average value of the demodulated signal used in the correction means and used in the determination means during the second period;
A receiver characterized by that.   The receiver according to claim 1, wherein the means for correcting the average value of the modulation signal uses a frequency correction loop.   3. The receiver according to claim 1, wherein the second extraction unit includes a low-pass filter unit that extracts an average frequency of the signal. 4.   2. The first estimating means includes means for evaluating a minimum value and a maximum value of the received signal and estimating an average value of the signal as an intermediate value between the minimum value and the maximum value. The receiver according to claim 3.   2. The received signal includes a synchronization and control unit and a subsequent data unit, and the first period does not exceed a period necessary for receiving the synchronization and control unit. The receiver of Claim 4 thru | or 4.   An integrated circuit comprising the receiver according to claim 1. A method for receiving and processing a signal received in a wireless network comprising:
A demodulation step of demodulating the received signal;
A step of estimating an average value of the demodulated signal;
A step of correcting the average value of the demodulated signal to a reference value;
Determining a binary value obtained by the received signal, and
The estimation step includes
A quick sub-step for extracting a first average value of the demodulated signal used in the determination step during a first period;
A second sub-process that is used in the correction step and that is used in the determination step and that is used in the determination step to extract a second average value of the demodulated signal at a low speed;
A method characterized by that.

Images